Recovering Explanations from Transformed Rule-Based Ontologies

2026-07-31

AI Summary

This work addresses the loss of interpretability in Datalog-based ontology reasoning caused by rule rewriting optimizations, which obscure the original rule structure and hinder the reconstruction of human-understandable inference traces from optimized proofs. We formally define the problem of reconstructing original-rule proofs from optimized ones, analyze its computational complexity, and introduce two practical proof transformation languages. Building upon Datalog rule rewriting, proof tree transformations, and complexity theory, we develop an efficient conversion mechanism that preserves logical equivalence while restoring explainability at the level of the original rules. This approach provides a theoretically grounded foundation for traceable and interpretable reasoning in knowledge graph applications.

Abstract

Datalog rules are often used to define ontologies over Knowledge Graphs. Rule reasoners routinely optimise such ontologies by rewriting their rules into a form that can be evaluated more efficiently. These transformations preserve the entailed facts, but not the structure of the underlying derivations. A proof tree under the rewritten rules explains why a fact holds, but does not readily yield an explanation in terms of the original rules. We study the problem of constructing, from a proof of entailment under the rewritten rules, a proof under the original ones: we establish its computational complexity and identify two practically relevant languages for specifying proof transformations.

Problem

• Datalog
• ontology
• rule rewriting
• proof transformation
• explanation recovery

Innovation

• Datalog
• ontology rewriting
• proof reconstruction
• explanation recovery
• rule-based reasoning